AI-Powered COVID-19 Detection from Chest X-Rays | Research Project
Medical Disclaimer: This tool is strictly for research and educational purposes only. It should not be used as a substitute for professional medical advice, diagnosis, or treatment. Always consult qualified healthcare professionals for medical guidance.
This project leverages deep learning technology to analyze chest X-ray images for potential COVID-19 indicators using the comprehensive COVID-19 Radiography Dataset. Built with PyTorch and featuring an intuitive Streamlit web interface, this tool demonstrates the potential of AI in medical image analysis while maintaining a focus on research and educational exploration.
The COVID-19 Radiography Dataset is a comprehensive collection of chest X-ray images created by researchers from Qatar University, University of Dhaka, and their collaborators. The dataset contains:
- 2,473 images from padchest dataset
- 183 images from German medical school
- 559 images from SIRM, Github, Kaggle & Twitter
- 400 images from additional Github sources
- 8,851 images from RSNA
- 1,341 images from Kaggle
- 6,012 Lung Opacity images (Non-COVID lung infection) from RSNA
- 1,345 Viral Pneumonia images from various sources
All images are in PNG format with 299x299 pixel resolution. The dataset is regularly updated with new X-ray images as they become available.
- Deep learning model for COVID-19 detection from X-rays
- Interactive web interface using Streamlit
- Real-time predictions with confidence scores
- Data augmentation and preprocessing pipeline
- Comprehensive error handling and logging
- Medical disclaimer for responsible use
.
├── app.py # Streamlit web application for model inference
├── data_preparing.ipynb # Jupyter notebook for data preprocessing
├── trainer.ipynb # Jupyter notebook for model training
├── model.pth # Trained PyTorch model weights
├── requirements.txt # Python package dependencies
├── LICENSE # MIT License file
├── README.md # Project documentation
├── .gitignore # Git ignore rules
├── assets/ # Static assets and resources
├── prepared_data/ # Directory for preprocessed dataset
├── prepared_data.zip # Compressed preprocessed dataset
└── .venv/ # Python virtual environment
app.py: Main Streamlit web application that provides the user interface for COVID-19 detectiondata_preparing.ipynb: Notebook containing data preprocessing pipeline, augmentation, and dataset preparationtrainer.ipynb: Notebook with model architecture, training loop, and evaluation metrics
model.pth: Pre-trained deep learning model weights (181MB)- Alternative download link: Download model weights
prepared_data/: Directory containing the preprocessed and augmented datasetprepared_data.zip: Compressed version of the preprocessed dataset (513MB)- Download link: Download preprocessed dataset
requirements.txt: List of Python package dependenciesREADME.md: Comprehensive project documentationLICENSE: MIT License file.gitignore: Specifies which files Git should ignore
assets/: Directory containing static resources like images and styles.venv/: Python virtual environment directory containing project dependencies
- Clone the repository:
git clone https://github.com/Sayemahamed/AI-Lab-Project.git
cd AI-Lab-Project- Install the required packages:
pip install -r requirements.txt- Start the Streamlit application:
streamlit run app.py-
Open your web browser and navigate to the provided URL (typically http://localhost:8501)
-
Upload a chest X-ray image and get real-time predictions
- Custom PyTorch Dataset implementation
- Advanced data augmentation techniques
- Stratified train/validation split
- Class weight calculation for imbalanced data
- Comprehensive error handling and logging
- Modern, responsive UI
- Real-time predictions
- Probability visualization
- Medical disclaimer
- Error handling and logging
- Support for various image formats
See requirements.txt for a complete list of dependencies. Key packages include:
- PyTorch and torchvision for deep learning
- Streamlit for web interface
- Pillow for image processing
- NumPy for numerical computations
- scikit-learn for data splitting
M.E.H. Chowdhury, T. Rahman, A. Khandakar, R. Mazhar, M.A. Kadir, Z.B. Mahbub, K.R. Islam, M.S. Khan, A. Iqbal, N. Al-Emadi, M.B.I. Reaz, M. T. Islam, "Can AI help in screening viral and COVID-19 pneumonia?" IEEE Access, vol. 8, pp. 132665-132676, 2020.
T. Rahman, A. Khandakar, Y. Qiblawey, A. Tahir, S. Kiranyaz, S.B.A. Kashem, M.T. Islam, S.A. Maadeed, S.M. Zughaier, M.S. Khan, M.E. Chowdhury, "Exploring the Effect of Image Enhancement Techniques on COVID-19 Detection using Chest X-ray Images." Computers in Biology and Medicine, vol. 132, pp. 104319, 2021.
The dataset includes images from multiple sources that should be acknowledged:
-
RSNA Pneumonia Detection Challenge Dataset
Wang X, Peng Y, Lu L, Lu Z, Bagheri M, Summers RM. ChestX-ray8: Hospital-scale Chest X-ray Database and Benchmarks on Weakly-Supervised Classification and Localization of Common Thorax Diseases. IEEE CVPR 2017.
-
PadChest Dataset
Bustos A, Pertusa A, Salinas JM, de la Iglesia-Vayá M. PadChest: A large chest x-ray image dataset with multi-label annotated reports. Medical Image Analysis, 2020.
If you use this project in your research, please cite it as:
@software{COVID19_XRay_Detection,
author = {Sayem Ahamed},
title = {COVID-19 Detection from Chest X-rays},
year = {2023},
publisher = {GitHub},
url = {https://github.com/Sayemahamed/AI-Lab-Project}
}- GitHub Repository: https://github.com/Sayemahamed/AI-Lab-Project.git
- Contact: sayemahamed183@gmail.com
This project is licensed under the MIT License.
Contributions are welcome! Please feel free to submit a Pull Request.